基于互信息法和改进模糊聚类的温度测点优化

发布时间：2018-05-21 14:44

本文选题：互信息法 + 改进模糊聚类　；参考：《仪器仪表学报》2015年11期

【摘要】：基于热误差模型进行机床热误差补偿是保证数控机床加工精度的一种有效方法,温度测点的布置和辨识会直接影响热误差建模的精确性和鲁棒性。本文提出一种互信息和改进模糊聚类法相结合的机床热关键点优化方法。以机床不同位置处的多个测点温度值及工件热变形作为分析数据,通过计算温度变量与热变形之间的平均互信息量,获得其综合关联度矩阵,确定二者之间的相关性后初选温度变量。根据改进模糊聚类法、F统计量和复判定系数对初选的温测点进行聚类,并结合温度变量与热变形之间的综合关联度值提取机床热关键点,从而实现测点优化。将基于该方法所得到的热误差模型与采用变量分组优化法获得的热误差模型进行比较,结果显示采用该方法进行热误差建模,机床X轴和Y轴的热变形预测精度得到显著提高,有利于改善加工精度。
[Abstract]:Thermal error compensation based on thermal error model is an effective method to ensure the machining accuracy of NC machine tools. The layout and identification of temperature measurement points will directly affect the accuracy and robustness of thermal error modeling. In this paper, a method for optimizing the thermal critical points of machine tools is presented, which combines mutual information with improved fuzzy clustering method. Taking the temperature values of several measuring points and the thermal deformation of the workpiece at different positions of the machine tool as the analytical data, the comprehensive correlation degree matrix is obtained by calculating the average mutual information between the temperature variables and the thermal deformation. Determine the correlation between the two after the primary temperature variable. According to the improved fuzzy clustering method and the complex decision coefficient, the primary temperature measuring points are clustered, and the key points of the machine tool are extracted by the comprehensive correlation value between the temperature variables and the thermal deformation, so as to realize the optimization of the measuring points. The thermal error model based on this method is compared with the thermal error model obtained by the method of variable grouping optimization. The results show that the thermal error prediction accuracy of the X and Y axis of the machine tool is improved remarkably by using this method to model the thermal error. It is propitious to improve the machining precision.
【作者单位】：西安理工大学教育部数控机床及机械制造装备集成重点试验室;陕西理工学院机械工程学院;
【基金】：国家自然科学基金(51375382) 陕西省自然科学基金(2013JM7014) 陕西省教育厅产业化培育(2013JC27)项目资助
【分类号】：TG659

【参考文献】